In a performance collection system, it often needs to periodically collect a plurality of kinds of performance items from a plurality of collection points of a device. The collection point refers to a physical entity on the device that can provide collection and statistics of performance data and is an appointed source on the device for obtaining performance data. These performance data will be stored in a database, and moreover are to be capable of generating data reports with a greater granularity periodically. Condition query will be often performed on the data in the reports to display the data in a report query Graphical User Interface (GUI), and overdue data in these reports are to be able of being deleted periodically.
The current storages of performance data are all done in the following way: one kind of performance items is stored in one database table, and the report data with the same generation granularity of the same kind of performance item are stored in one database table. When the collection scale is very large, that is, when the number of collection points is too big, a larger amount of processing on performance data will be involved, including storage of these data, deletion of overdue data and condition query of these data. Because the data quantity is too big, when a data table stores a large volume of data, basic operations such as Create, Update, Read, Delete (CURD) will all affect each other, thereby reducing the processing efficiency and influencing the availability of the system itself.